Language-Specific Model Versioning Based on Signifiers.

Abstract

In model-driven engineering (MDE), models constitute the central artifacts in the development process, and thus, are often built by teams of developers. As a consequence, adequate version control for models is crucial to the success of MDE-based projects. Several model versioning systems have been proposed recently. Most of them are generic in the sense that they are agnostic to modeling languages. Although this ensures a wide applicability, important merge issues may not be detected.In this paper, we present an orthogonal extension to generic model versioning systems for enabling the detection of an important subset of language-specific merge issues. Users may enhance the versioning system's capabilities by defining signifiers, which describe the combination of features of a model element type that convey the superior meaning of its instances. Signifiers improve the different phases of the versioning process including comparing and merging models leading to a higher quality of the finally merged models. We showcase the applicability of our approach by enhancing the versioning support for the modeling language Ecore.